572 THE AMERICAN NATURALIST. [Vor. XXXV. 
of regenerating new somites in case of their accidental loss. 
Syllis ramosa, a remarkable annelid described by M'Intosh 
(85), is not only capable of forming terminal somites, but 
also of producing lateral buds. The branches of the main 
axis thus formed may again bud, forming a complex system, 
any ultimate tip of which may be constricted off and develop 
into a sexual animal. In this case a portion only of a somite 
may give rise to a perfect individual. As Andrews has pointed 
out, lateral budding in Syllis, and bifurcation of the posterior 
end in Lumbricus, are undoubtedly merely different manifesta- 
tions of the same phenomenon. Inthe budding, duplications of 
the main axis are formed at successive periods ; in bifurcations of 
the body, the duplications are formed simultaneously. In both 
cases a sexual animal exhibits the power of reproducing itself 
by means of a vegetative process. 
It is easily conceivable that, as a result either of injury to 
the anal somite, or loss of caudal metameres, two regions 
of regeneration might be formed instead of one. If these 
formative areas are entirely independent of each other, their 
continued growth would produce a bifurcated condition, and 
Bülow (83) occasionally found this to be the case in regenerat- 
ing earthworms. If, however, the separation of the two areas 
of regeneration is incomplete, the duplication of the new meta- 
meres would be only partial. Thus, in the Nereis which we 
have described, injury to the anal somite (when it was the 
forty-first metamere of the worm), or accidental loss of a few 
somites posterior to the fortieth, might, on regeneration, have 
produced the incomplete duplication of the metameres. That 
this may be the true explanation of the abnormality seems very 
probable, but further (experimental) evidence is needed to 
demonstrate it beyond a doubt. 
It would seem a more difficult task to account for the gradual 
diminution in size and final disappearance of the interpolated 
parts. The changes which brought about such a return to the 
normal condition could have taken place only in the anal meta- 
mere, for it is there that the fundament of each new jonin 
is laid down. The gradual process of reconstruction which 
must have taken place in the formative tissue may be compared 
